Teachers' Manual to Green, Nesson & Murray, Problems, Cases and Materials on Evidence, 3rd Edition.

Go to: Teacher's Manual Introduction Chapter 8 Contents Next Section

C. Expert Testimony, Scientific Proof, and Junk Science (771)

Frye v. United States, 293 F. 1013 (D.C. Cir. 1923) (772)

The Frye court is responding to a basic problem posed by expert testimony in novel areas: the judges who must decide on the admissibility of the evidence (FRE 104(a)) are incapable of directly judging the efficacy of the novel approach, and so must develop some mode of judging it based on secondary indicators.

The Frye court chose an approach which responded to the central judicial objective of producing judicial verdicts which could be credited and respected by the general public. The court therefore chose a method which would assure some general acceptance and deference to the methods of proof on which verdicts could be based. The question underlying the court's verbal formulation is "would a verdict based on the outcome of the new scientific method be respected and accepted." The courts answer is, in effect, "only if people who are knowledgeable and respected would accept it." The objective of the Frye test, in other words, is to provide a method by which lay judges can assess the potential acceptability by a lay public of a scientific process which neither the judges nor the public have the knowledge to evaluate directly.

Frve, looked at more narrowly as a case that rejected a lie-detector expert, could be understood as turning on the newness and lack of acceptance of the lie-detector test at the time (1923) when the case was decided. That would suggest that with fifty years of subsequent development of lie detection methods since Frye and with the wide acceptance of lie-detection as a useful tool in interrogation, lie Detector tests would now be readily admissible as evidence in courts. Such is not the case. Judicial resistance to the admissibility of lie detector tests has continued with little exception. See Commonwealth v. Lukus, 367 Mass. 191, 327 N.E.2d 671 (1975).

One suggested reason for this continued resistance is that lie detection is not infallibly accurate. While this concern about lie detection is wellfounded, it hardly explains the resistance to admissibility. The admissibility of other evidence is not conditioned on infallibility.

The problem with lie detection is that it threatens to supplant rather than enrich the factfinding process. In cases in which polygraph evidence is introduced, the jury may be tempted to base its verdict solely on the evidence of the test. In such cases, the acceptability of verdicts based on liedetection is no greater than the acceptability of the accuracy of lie-detection. This is what leads to the instinct that lie-detection would have to be 100% accurate to be admissible. The issue, posed another way, is not whether lie detectors would produce a higher percentage of accurate verdicts than does the trial process without lie-detectors. The issue is whether the lie detector would produce more acceptable verdicts. The question whether reliability and acceptability are the same thing, and if not, which should be the primary objective of judicial factfinding when they are in conflict, is one of the most profound questions of the evidence course, and is, obviously, very well posed by the issue of expert scientific proof. See generally, IA, Wigmore on Evidence sec. 37.1 at pp. 1018-1019 (Tillers Rev. 1983) characterizing the focus on acceptability as opposed to rationality as a "deviant legal theory."

This brings a further question into focus. It is one thing to insist on acceptable verdicts of guilt, and thus to require the prosecution to use only tried and true, generally accepted scientific methods of proof. It is quite another matter to insist that defendants use only such generally accepted methods in seeking acquittals. To impose such a requirement on defendants seriously undercuts the acceptability of convictions in cases in which the defendant claimed to have reliable but not generally accepted scientific evidence to prove his innocence.

As if to underscore this point, Frye was exonerated some years after his conviction when another person confessed to the murder. Wicker, The Polygraphic Truth Test and the Law of Evidence, 22 Tn.L.R. 711, 715 (1953).

Perjury Case Goes to the Dogs, National Law Journal, 9/14/80 (775)

Reversal in this case promoted the following values:

1. Criminal prosecution is supposed to be a serious business. Making too much of a joke out of a case will get it reversed.

2. Although there is a patina of plausibility to the idea that an expert could determine the date of a photograph from the length of the shadows in the picture, there is no developed practice of doing this. Prosecution is a conservative business which should rest on "tried and true" methods of proof.

3. Even if it is possible to determine dates from shadows, there is no developed understanding of the difficulties of doing this; the cross-examiner is therefore at a considerable disadvantage, and there will be great difficulty in locating other experts.

We cannot tell from the National Law Journal write-up exactly what theory the astronomer applied, or how carefully he explained it. Shadow length is not unique to specific days (shadows range from long to short to long every day), but the combination of shadow length and shadow direction may be. Had the case been more important than a minor perjury prosecution against someone who tried to help his brother, and had the case been handled with more seriousness, it is perfectly possible that the astronomer's approach could have been approved.

Problem - Frye and the Defense (776)

Answer and Analysis:

The question presented by this hypothetical problem is whether the defendant's constitutional right to offer relevant evidence in his defense should apply so as to require a court to admit the defendant's polygraph evidence. Does the constitutional guarantee of compulsory process lower the threshold for admissibility of expert testimony proffered by a criminal defendant?

Problem - The Dowser (777)

The point:

General acceptance of the method of proof determines general acceptance of the verdict. Therefore, if acceptance of the system's verdicts is an objective of the system of judicial factfinding, then that objective dictates that methods of proof be judged according to their general acceptability.

Answer and analysis:

The testimony is inadmissible, unless dowsing is so well accepted in the community in which the deal at issue in this case was made that the contract could be interpreted as if it said "unless no water can be found within 400 ft. of the surface as indicated by dowsing."

Absent such a strong local acceptance of dowsing, the testimony would be inadmissible because dowsing is not a generally accepted scientific method of determining the presence of water. This is not to say that it is not an effective method. A court need take no position on this. It is only to say that dowsing is not generally accepted as such. General acceptance of the method is important because general acceptance of verdicts determined by the method depends upon it.

Problem - Good Buddy (777)

The point:

Whether a witness can qualify as an expert depends on the relative specialization of his knowledge and experience vis-a-vis the factfinders. The question is whether he can be helpful to these factfinders in this case.

Answer and analysis:

The judge should decide to rule under FRE 702 because jurors could not all be expected to have sufficient familiarity with trucks to gauge the severity of bumps in terms of their capacity to cause loads to shift. Under FRE 702 even though trucking might not normally be considered a type of expertise, the witness could qualify to testify to his opinion about the severity of the bump because of his experience.

It matters procedurally that the question is considered under FRE 702 rather than FRE 701 because under FRE 702 the defendant would be entitled to contest the qualifications of the witness on voir dire.

Problem - Abortion or Manslaughter (777)

The point:

The party offering an expert witness to testify to his opinions must establish the foundational elements for such testimony to the satisfaction of the judge pursuant to FRE 104(a).

Answer and analysis:

D is entitled to cross-examine W about his qualifications. W may not give his opinion unless he has qualified as an expert under FRE 702. Whether he has or not is a decision to be made by the judge under FRE 104(a). The judge must determine whether the witness is qualified to testify, not merely whether a jury could find him to be qualified. Therefore D is entitled to have the judge hear both sides before he decides: this means that D is entitled not only to cross-examine W about his qualifications but also to call his own witnesses to testify to W's absence of qualifications.

United States v. Johnson, 575 F.2d 1347 (5th Cir. 1978) (777)

The opinion in this case turns the approach of the Frve test upside down. By some magic the witness purported to be able to identify the source or origin of marijuana. The court permitted this because "we cannot say that the claim of an ability to identify Colombian marijuana is so inherently implausible that, as a matter of law, a jury should not be permitted to hear testimony on the identification." On this approach accidentologists, dowsers, and astrologists would be allowed to testify, to say nothing of lie detector analysts, voiceprint analysts, and experts about eyewitness identification.

Moreover, because the witness's testimony was the only evidence that the marijuana came from outside the customs territory of the United States, the court was necessarily holding that the expert's testimony was sufficient to support a finding of importation.

One explanation for this case is that the Fifth Circuit is winking at the rules with respect to experts order to take the content out of the statutory requirement that the government must prove importation as an element of its case. The reality is that the jury, if given the chance, will not acquit a major marijuana dealer on such technicality, and the Fifth Circuit is ensuring that the jury will get the chance to be sensible. A more charitable view is that the court is applying the test of the Federal Rules rather than the Erie test, and is persuaded that the "expert" has sufficient specialized knowledge based on his experience with marijuana to make his testimony helpful. By any standard, however, the rule is not that a witness may testify as an expert if his claim to expertise is anything better than "inherently implausible."

Questions at page 782

The proposed alternative differs from FRE 703 by making admissible that which is "reasonably and customarily relied upon" even if it is hearsay. Under FRE 703 it is not admissible. FRE 703 permits the expert to itly on the hearsay, and he may reveal the facts on which he has relied to the jury, but these hearsay facts are revealed only for purpose of showing the basis for the expert's opinion, and not for the truth of the matter asserted. The opposing party is entitled to a limiting instruction to this effect. The distinction is virtually impossible to grasp, therefore its maintenance and its articulation in instructions to the factfinder can only produce confusion. The alternative proposal is preferable because it eliminates a distinction which is untenable, and because it forces the proponent of the expert to bring more of the evidence on which the expert has relied to court.

Problem - Engineer's Investigation (782)

The point:

The concept of reasonableness incorporated into FRE 703 may introduce a form of analysis very similar to that applied in evaluating hearsay exceptions.

Answers and analysis:

(1) The statement by the driver reported to the engineer is hearsay. The problem is hardest if it is assumed that the driver's statement is not otherwise admissible and has not otherwise been made known to the jury.

The procedural structure within which this problem must be analyzed is clear, even if its outcome is doubtful. The issue of the admissibility of the expert's opinion is a question to be decided persuant to FRE 104(a). A foundational element in establishing admissibility is to show that the bases for the opinion are permissible. Therefore P bears the burden of convincing the judge that engineers reasonably and customarily rely on hearsay statements by drivers in deciding whether accidents are caused by design defects.

We do not know whether or not engineers would rely on such hearsay. Certainly engineers trying to determine the cause of aircraft accidents rely on recorded statements by the crew, but then such statements are contemporaneous to the event (admissible under FRE 803(1)) and made by highly trained personnel. This suggests that engineers would probably rely on reliable hearsay statements, but probably would not rely on a statement made by the driver a week after the accident after being advised by his lawyer that the best chance of recovery in a lawsuit would be based on a claim of design defect in the steering column. In other words, engineers probably size up the reliability of hearsay much as judges or anyone else would. This suggests that the requirement that the engineer "reasonably" rely on the hearsay may come close to superimposing on FRE 703 the various considerations which are reflected in the hearsay exceptions. A voir dire on reasonable reliance would be necessary and the judge would have to decide.

(2) Yes, the engineer could state her opinion without revealing the basis for it. FRE 705. Opposing counsel can inquire what the basis for the opinion is, but takes some risk of strengthening the expert's presentation in doing so. With adequate pretrial discovery the cross-examiner might know in advance that the expert based her opinion in part on a hearsay statement by the driver. In such a case the cross-examiner could object to the admissibility of the opinion and ask to examine the expert out of the presence of the jury in order to demonstrate that the opinion is based on impermissible grounds. Under the Federal Rules of Civil Procedure, discovery of expert opinion, while limited, is adequate to this task. Not so under many state rules of procedure, which makes the adoption of FRE 705 by these states problematic.

(3) The case would be different because pretrial discovery in criminal cases is much more limited than in civil cases. Counsel would not be able to assert any basis for thinking that the opinion was based on impermissible grounds, and would, therefore be fishing blindly.

Problem - Pediatric Poison (782)

The point:

The combined effect of the federal opinion and hearsay rules has been to greatly ease the difficulties of using learned treatises in the direct and cross-examination of experts.

Answers and analysis:

(1) Section 904 of Taylor's Medical Jurisprudence is hearsay, hence inadmissible, at common law.

(2) While there may have been a question as to the proper outcome in this situation at common law, it is clear under the federal rules that the objection should overruled. There are two separate reasons.

First, FRE 703 permits an expert to base his opinion on information which need not itself be admissible, if of a type reasonably relied upon by experts 'in the particular field in forming opinions upon the subject. Certainly an established text would be a source of information typically and reasonably relied upon by experts in the field.

Second, under FRE 803(18) the treatise would be admissible under the learned treatise exception to the hearsay rule because it was relied by Dr. Kildare on direct examination and could be established as a reliable authority by Dr. Kildare.

At common law, the rules for using treatises were much more restrictive. At their narrowest, a witness could not be cross-examined on the basis of a treatise unless he had himself relied upon it on direct, or unless he himself recognized the treatise as authoritative in the field. This meant that a cross-examiner could be very substantially constrained if the witness simply refused to recognize the authority of texts that contradicted him. This problem is explored in the film, Edelin Eleventh Day.

State v. Saldana, 324 N.W.2d 227 (Minn. 1982) (784)

This case addresses the issue of overlap between expert testimony and general experience. The decision reflects a reluctance to consign to expert testimony matters which have traditionally been considered and weighed directly by the jury. How is the jury to consider and weigh the expert's opinions in comparison with their own experientially based inferences? Yet as ongoing progress in science and specialized learning provides insights beyond those afforded by everyday experience, shouldn't this specialized information be made available in the trial of cases?

The dilemma faced by the court in Saldana with respect to psychological expert testimony bears some resemblance to that faced by the court in with respect to statistical and probabilistic expert testimony. In both cases the subject matter of the expert opinion overlapped with general human experience. In both cases the court was concerned that the jury would have difficulty evaluating the expert testimony and comparing it to inferences borne of experience. In both cases the court's decision was eased by apparent defects in the methodology of the expert. We now know that where the science is solid, both statistically based (e.g. DNA) expert evidence, and psychological (see, e.g. Chapple, 789) expert testimony are admitted to be considered by the jury along with its pool of real-life experience.

State v. Chapple - 135 Ariz. 281, 660 P.2d 1208 (1983)(789)

This case is an early example approving the admission of psychological expert testimony on an issue (the validity of eyewitness identification) previously considered as the province of the jury based on common experience. One factor supporting admission was the fact that the testimony was offered by the defendant. The policy behind affording the accused every opportunity to adduce proof of innocence may have been the extra weight in the scales. Since Chapple many courts have come to approve admission of expert testimony on the reliability of various forms of inculpatory testimony. Elizabeth Loftus, the witness involved in Chapple, has become a national expert on the subject and has testified in courts across the nation.

Generally the defendant has had an easier time securing admission of overlapping expert testimony than the prosecution. Prosecution expert testimony that the defendant possessed the psychological profile of a serial killer or a pedophile when offered as evidence of guilt would still be regarded with suspicion. Expert testimony from either prosecution or defense which associates behavior-influencing psychological conditions with crime victims (rape trauma syndrome, abused spouse syndrome) or with alternative suspects (abused child syndrome - see Conlogue, Chapter III above), is regularly received.

State v. Williams, 388 A. 2d 500 (Me. 1978) (799)

This is an early state supreme court case considering the effect of a state analogue to FRE 702 on the minimum standard of quality required to sustain the admission of expert testimony. The question of whether the Frye test was repealed by the language of Rule 702 was not answered by the United States Supreme Court until 20 years later in Daubert v. Merrell Dow (below).

The Maine court's opinion in Williams construes Maine Rule 702 to require a flexible relevance-based approach to determining admissibility of expert testimony. If the expert testimony is not of a given minimum quality it is not relevant - i.e., the court cannot find that it makes a fact of consequence to the outcome of the action more or less likely. The court thus screens the quality of the science pursuant to its duty to admit only evidence that makes a difference.

The simplicity and elegance of the approach of the Maine Supreme Court in Williams can be contrasted with the approach of the United States Supreme Court in Daubert below. After considering the trilogy of United States Supreme Court decisions on expert testimony (Daubert, Joiner and Kumho) aren't we finally arrived at the same place the Maine Supreme Court was in 1978?

Barefoot v. Estelle, 463 U.S. 880 (1983)

Is there a constitutional "bottom line" to the admissibility of expert testimony in criminal cases? Does a person accused of crime have a constitutional due process right to a minimum threshold of reliability for expert testimony offered by the prosecution?

In Barefoot v. Estelle highly questionable psychiatric predictions of the defendant's future dangerousness were admitted at the penalty phase of a capital murder prosecution. Defendant's contention that the Due Process clause establishes a floor for the admissibility of such evidence was given short shrift by the majority. While the court has not hesitated to regulate lay testimony in the interest of due process, Barefoot represents a reluctance to question standards of admissibility of expert testimony established in the states.

In re Agent Orange Product Liability Litigation (Lilley), 611 F. Supp 1267

(E.D.N.Y. 1985) (822)

This case illustrates the determination of admissibility of expert testimony at the summary judgment stage of a civil case. This practice is of increasing significance in cases of medical malpractice, toxic torts, and products liability, where expert testimony is deemed necessary to get a case to a jury. If the expert testimony or the expert is excluded, the case is over.

One cannot help but get the impression that the court is engaging in considerable fact finding about the expert and what the expert did to arrive at the challenged opinion when it evaluates the proffered expert testimony. Such evaluations are generally the province of the jury in the case of lay testimony. How far should the court go in analyzing and accrediting or discrediting expert testimony in the decision of whether to admit?

Wells by Maihafer v. Ortho Pharmaceutical Corp., 615 F. Supp. 262 (N.D. Ga. 1985), modified, 788 F.2d 741 (11th Cir. 1986) (841)

This case illustrates the difficulty encountered even by a competent judge in choosing between competing expert opinions. Without an independent knowledge and grounding in the science in question how is the judge to know which expert opinion to accept. Our adversary system often produces diametrically opposed expert conclusions from highly credible exponents. What is a judge to do?

Judge Shoob was commendably frank in his attempt to explain and justify the rationale by which he decided to accept the plaintiff's experts in preference to those of the defendant. What he did not foresee was how this rationale would strike third-party observers.

Federal Judges v. Science, N.Y. Times, Dec. 27, 1986, p.22, col. 1 (848)

This reaction to Judge Shoob's opinion in Wells v. Maihafer, above, is not surprising. Should weighty disputes of scientific causation be decided on the basis of the demeanor and teaching ability of the exponents of the competing opinions? The problem is that there is no simple solution to the problem represented by this case. How is a judge who originally went to law school because he hated science supposed to decide highly disputed questions of sophisticated science? How about a jury?

Daubert v. Merrell Dow Pharmaceuticals, Inc. , 113 S. Ct. 2786 (1993)(849)

The Daubert case is so well known to Evidence teachers that there is no need for commentary on it. Most teachers would spend some time on the following issues:

1) How is the judge supposed to evaluate this "science" as envisioned by the Supreme Court? Can a judge, for example, rely on a ruling made by another judge looking at the same or a similar proposition

2) What about other kinds of expertise that would not be considered "science"? Should such expertise be subjected to a threshold inquiry before submission to the jury?

    1. How does this decision affect the judge-jury division of roles? Does this permit the judge to invade the jury's responsibility to find the relevant facts?

Some, but not all, of these question are answered by cases following Daubert.

Daubert v. Merrell Dow Pharmaceuticals (on remand), 43 F. 3d 1311 (1995) (860)

This case is an early example of application of the Daubert in the summary judgment context. Complaining about the role thrust upon them by the Supreme Court, the judges of the 9th Circuit alternately apply the two "legs" of Daubert to exclude all of the plaintiffs' key expert testimony and hence defeat the case. The court concludes that the methodology of the one doctor who testified to a positive causal link between the defendant's products and the plaintiff's injuries was scientifically insufficient to pass the Daubert "scientific reliability" test. The court then finds that the testimony of the plaintiff's other doctors, concededly based on sound science, did not go far enough to "fit" the issue to be decided by the jury. The exclusion of all of the witnesses' testimony on one or the other theory resulted in summary judgment for the defendant.

While the process of the 9th Circuit seems logically sound, one wonders whether such a witness-by-witness analysis ignores possible synergies among the witnesses which would make their overall testimony of greater value or credibility than the testimony of any single witness. Should the issue be one of admissibility of a single witness's testimony, or should it be considered one of sufficiency of the scientific proof as a whole? Does our fear of the jury's "black box" reaction cause us to shoehorn appraisals of evidentiary sufficiency into determinations of admissibility?

Glaser v. Thompson Medical Co., 32 F.3d 969 (6th Cir. 1994) (869)

This case represents another early example of Daubert in practice. The court finds that the position of plaintiff's expert, though controversial, has some sound "scientifc" backing. Accordingly, the court denies the motion for summary judgment allowing the case to proceed to trial by jury. The court's confident suggestion that the dispute among scientists about the role, if any, of Dextramin an causing hypertension, can be resolved by the jury gives rise to the question, "How is the jury supposed to figure out which of the disputing experts is right?"

Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999) (873)

In this case the Supreme Court applied a Daubert-like threshold analytical process to expert testimony not fitting the description as "scientific". Daubert had relied on the connotation of the word "scientific" in inferring a minimum level of regularity and reliability for the kind of expert testimony before the court in that case. What defines the threshold for expert testimony of a non-scientific nature? The analogous kind of inquiry required by Kumho for non-scientific expert testimony may work well for engineering and other technical testimony. But how about expert testimony based on specialized familiarity of a non-technical nature? See Problem - Good Buddy, above. Would we be better off with a flexible relevancy based inquiry (Williams?) than we are with an inquiry somehow tied to science and technology and the methods of these disciplines?

General Electric Co. v. Joiner, 522 U.S. 136 (1997) (886)

This case raises the question of the standard of review of a trial court's exclusion of proffered expert testimony in the context of a summary judgment. The Court of Appeals was troubled by the fact that in the context of a summary judgment, the exclusion of proffered expert testimony can be "outcome determinative" of the entire case. Should this require closer scrutiny of the decision on appeal? "No", said the Supreme Court, ruling that a decision to admit or exclude expert testimony is like any other relevance determination, reviewable only for abuse of discretion.

An abuse of discretion standard makes the most sense when one is considering rulings made by a trial judge during the course of trial. The trial judge has the complete and immediate view of the proceedings as a whole and is in the best position to weigh all competing considerations. An appellate court should be reluctant to second-guess her calls except where they are clearly out of bounds.

On the other hand, one cannot avoid feeling a bit uneasy granting the same degree of deference to isolated, briefed and argued determinations incident to motions to summary judgment. Such determinations have the effect of squeezing issues of sufficiency into issues of admissibility, and of reaching final case outcomes based on weighing and analysis of individual witness's testimony rather than the evidentiary record as a whole. Would it make more sense to subject at least exclusions of expert testimony which result in case dismissal to somewhat greater appellate examination?

United States of America v. John Ray Bonds, et al., 12 F. 3d 540 (6th Cir. 1993) (896)

This is an example of application of Daubert to the then-new science of DNA testing. The court attempts to distinguish between questions about the reliability of the science, which are to be resolved by the court as an precondition of admissibility, and questions about the reliability of the specific test, as to which there is some play for the jury. This case offers another good opportunity for discussion of the effect of DNA evidence on the jury's fact-finding function.

Go to: Teacher's Manual Introduction Chapter 8 Contents Next Section